RESEARCH Open Access

Association between tea and coffeeconsumption and brain cancer risk:an updated meta-analysisYang Song1, Zhiyun Wang1*, Yanyu Jin2 and Jie Guo1

Abstract

Background: Previous studies had demonstrated some associations between coffee and tea consumption andbrain cancer risk resulted in an inconsistent relationship. We therefore performed this study to further explore theassociation between them.

Method: By searching PubMed, Embase, and Web of Science, we retrieved up to 1 November 2018, 11 relevantliterature of publications were collected by 2 people eventually. Stata 14.0 software was used for data analysis.

Results: In total, 11 articles (11 articles for coffee, 8 articles for tea, and 4 articles for coffee plus tea) were used inthis meta-analysis. A statistically significant protective effect of coffee consumption and brain cancer risk was found(RR = 0.785, 95% CI = 0.580–0.984, I2 = 65.6%, P for heterogeneity = 0.001), especially in Asian populations (RR = 0.217,95% CI = 0.042–0.896). However, the association between the risk of brain cancer and tea consumption was non-significant in the whole result (RR = 0.897, 95% CI = 0.739–1.088, I2 = 29.9%, P for heterogeneity = 0.189), but significantin American populations (RR = 0.798, 95% CI = 0.646–0.986). Interestingly, the RR was 0.684 (95% CI = 0.481–0.975)for the risk of brain cancer when compared the highest versus the lowest category consumption of coffee plus tea.

Conclusion: Findings from this study suggested that higher consumption of coffee may contribute to the lowerdevelopment of brain cancer in Asian populations. Tea consumption had an inverse association for the risk of braincancer in American populations, instead of other populations.

Keywords: Coffee, Tea, Brain cancer, Glioma, Meta-analysis

IntroductionBrain cancer is one of the most common malignant tu-mors in adult primary intracranial tumors [1, 2]. Therole of several nutritional factors in the prevention ofbrain cancer has been investigated. For example, con-sumption of fresh fruits and vegetables [3], fresh fish [4],and vitamin supplements [5–7] have been acted prevent-ive agents. Coffee and tea are popular with people astwo kinds of drinks. Coffee consumption had beenlinked to some cancers. Horisaki et al. performed adose-response meta-analysis indicated that coffee con-sumption may be weakly inversely associated with therisk of colorectal cancer in Japanese [8]. Previous studies

published by Lukic et al. [9] and Lafranconi et al. [10]suggested that a prospective effect of coffee consump-tion on the risk of endometrial cancer. Highest versuslowest category of tea consumption had been confirmedof decreased with liver cancer [11, 12], ovarian cancer[13], gastric cancer [14], oral cancer [15, 16], and endo-metrial cancer [17].A previous meta-analysis was published to explore the

association between coffee and tea consumption and therisk of glioma in adults [18]. They concluded that thehighest versus lowest category of coffee consumption,tea consumption, and coffee plus tea consumption hadno effect on the risk of brain cancer. However, manystudies had been published to further assess the associ-ation between coffee and tea consumption and the riskof brain cancer since that meta-analysis. For this reason,this paper increased the sample size and improved the

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

* Correspondence: wang__zhy@yeah.net1Department of Neurology, Tianjin First Center Hospital, No. 24, Fukang Road,Nankai District, Tianjin 300192, ChinaFull list of author information is available at the end of the article

Song et al. World Journal of Surgical Oncology (2019) 17:51 https://doi.org/10.1186/s12957-019-1591-y

test efficiency through an updated meta-analysis to ob-tain more authentic and reliable analysis results, whichis helpful to clarify whether coffee or tea consumptionhas some inverse effects on brain cancer development,and finally provides evidence of prevention for braincancer.

MethodsSearch strategy and inclusion criteriaTwo authors independently checked the databases ofPubMed, Embase, and Web of Science to find the rele-vant articles which investigated the association betweentea and/or coffee consumption and risk of brain cancerfrom inception up to 1 November 2018. The followingsearch terms were used: “tea” OR “coffee” OR “drinking”combined with “brain cancer” OR “brain tumor” OR“glioma.” To identify titles and abstracts of relevant lit-erature, reference lists of studies were checked manually.All studies included in this meta-analysis should meet

the following criteria: (1) assessment of the tea and/orcoffee consumption and brain cancer or glioma risk; (2)observational studies; (3) the detailed data of studiesmust be completely provided in all participants directlyor indirectly; (4) human studies.

Data extractionData were extracted independently by two researchers(YS and ZYW). All the information listed in Table 1 ineach study was abstracted carefully. The disagreementswith these two researchers were resolved by discussionand consensus.

Statistical analysisStata 14.0 software was used for data analysis. The com-bined relative risk (RR) with its 95% confidence interval(CI) for tea and coffee consumption and brain cancerrisk were calculated. Forest plots for the association be-tween tea and coffee consumption and brain cancer riskwere mapping. Heterogeneity between studies was evalu-ated by chi-square-based Q test and I2 test [19, 20]. Therandom-effects model was used for all the analysis. Po-tential publication biases were examined using Egger’slinear regression [21] and Begg’s funnel plots [22]. Sensi-tivity analysis was done to estimate the stability of theresults by removing a single study from the analysis.Subgroup analyses by study design and geographic loca-tion were also performed. P < 0.05 was considered statis-tically significant.

ResultsStudy selection and study characterizationAccording to retrieval methods above, 2421 relevantstudies from the databases (671 articles from PubMed,1299 articles from Web of Science, and 451 articles from

Embase) and 2 additional records identified when wereviewed the discussion and introduction of the includedstudies were selected. In total, 11 articles [23–33] wereincluded in this study, as shown in Fig. 1. Six studieswere from the USA, 1 from Australia, 1 from Canada, 1from Iran, 1 from Japan, and 1 from Europe (France,Italy, Spain, Denmark, Germany, Greece, Netherlands,Norway, Sweden, and the UK). Eight of the 11 includedstudies were cohort design, and the remaining 3 studieswere case-control design. The characteristics of the ob-servational studies are shown in Table 1.

Coffee consumption and the risk of brain cancerAll of the 11 included studies were used to assess the asso-ciation between coffee consumption and brain cancer risk.Pooled RR suggested that highest category of coffee con-sumption could reduce the risk of brain cancer (RR =0.785, 95% CI = 0.580–0.984, I2 = 65.6%, P for heterogeneity =0.001) (Fig. 2), when compared with the lowest category.Ten of the 11 included studies were performed about cof-fee consumption and glioma risk, resulting in a statisti-cally significant protective effect of coffee consumption onthe risk of glioma (RR = 0.760, 95% CI = 0.548–0.972).This result is showing a statistically significant protectiveeffect of coffee consumption in the cohort studies (RR =0.858, 95% CI = 0.700–0.992), instead of case-control stud-ies. Furthermore, we also explored whether the risk ofbrain cancer was related to geographic location. The re-sults from our analysis indicated that an inverse associ-ation was only found in Asian populations (RR = 0.217,95% CI = 0.042–0.896), but not in other populations else.Table 2 shows the results for both whole and subgroupanalyses.Based on Egger’s test (P = 0.214) and funnel plot

(Additional file 1: Figure S1), there existed no publica-tion bias. Sensitivity analysis was done, and the pooledRR ranged from 0.738 (95% CI = 0.542–0.961) to 0.905(95% CI = 0.754–1.088).

Tea consumption and the risk of brain cancerEight articles [24, 25, 27, 29–33] were included to ex-plore the relationship between tea consumption andbrain cancer risk. Our meta-analysis result showed thatthere was a non-significant association between tea con-sumption and brain cancer risk (RR = 0.897, 95% CI =0.739–1.088, I2 = 29.9%, P for heterogeneity = 0.189) (Fig. 3).Seven of the eight included studies were about tea con-sumption and glioma risk. The result was consistentwith the brain cancer (RR = 0.846, 95% CI = 0.683–1.047). We conducted a subgroup analysis by study de-sign (cohort studies and case-control studies). The asso-ciation was non-significant in prospective studies or incase-control studies. However, when we further exploredthe association between brain cancer and geographic

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 2 of 9

Table

1Characteristicsof

theinclud

edstud

ieson

teaandcoffeeconsum

ptionandbraincancer

risk

Stud

y,year

Cou

ntry

Stud

yde

sign

Participants

(cases)

Age

(years)

Outcome

Expo

sure

andcatego

ryRR

(95%

CI)

Adjustm

entforcovariates

Bagliettoet

al.2011

Australia

Coh

ort

39,766

(67)

27–81

Brainglioma

Coffee:4cups/day

ormore

vs.<

1cup

/day

0.51

(0.20–1.13)

Adjustedforsex,coun

tryof

birth,total

energy

intake

from

diet,and

levelo

fed

ucation.

Burchet

al.1987

Canada

HBC

C475(247)

25–80

Braincancer

Coffee:ever

vs.never

Tea:ever

vs.never

Coffee:1.26

(0.76–2.09)

Tea:1.13

(0.73–1.72)

Adjustedforage,sex,area

ofreside

nce,

maritalstatus,anddate

ofdiagno

sisor

death.

Dub

row

etal.2012

USA

Coh

ort

545,771(904)

50–71

Brainglioma

Coffee:≥6cups/day

vs.non

eTea:>3cup/dayvs.non

eTeaplus

coffee:>5cup/day

vs.non

e

Coffee:1.04

(0.70–1.55)

Tea:0.75

(0.57–1.00)

Teaplus

coffee:

0.68

(0.46–1.03)

Adjustedforage,sex,race/ethnicity,

height,and

intake

oftotalene

rgy,fru

it,vege

table,andnitrites.

Efird

etal.2004

USA

Coh

ort

133,811(130)

≥25

Brainglioma

Coffee:≥7cups/day

vs.<

1cup/day

1.7(0.8–3.6)

Adjustedforage,sex,race,smoking,

education,andalcoho

l.

Hashibe

etal.2015

USA

Coh

ort

97,334

(103)

55–74

Brainglioma

Coffee:≥2cups/day

vs.<

1cup/day

Tea:≥1cup/dayvs.<

1cup/day

Coffee:0.76

(0.50–1.17)

Tea:1.04

(0.65–1.66)

Adjustedforage(con

tinuo

us),sex,race,

anded

ucation.

Hochb

erget

al.1990

USA

PBCC

288(160)

15–81

Brainglioma

Coffee:≥4cups/day

vs.<

1cup/day

0.9(0.5–1.8)

Adjustedforage,sex,and

socio-econ

omic

status.

Holicket

al.2010

USA

Coh

ort

230,655(335)

25–75

Brainglioma

Coffee:≥4cups/day

vs.non

eTea:≥8cups/day

vs.non

eTeaplus

coffee:≥5cups/day

vs.<

1cup/day

Coffee:0.80

(0.54–1.17)

Tea:0.71

(0.45–1.12)

Teaplus

coffee:

0.60

(0.41–0.87)

Adjustedforage,sex,totalcaloricintake.

(Further

adjustmen

tsforcigarette

smoking,

curren

tsm

oking,

intake

ofprocessedmeat,alcoho

l,fru

itand

vege

tables

,and

,for

wom

en,rep

rodu

ctive

factorsdidno

tchange

theriskestim

ates)

Malmiret

al.2017

Iran

HBC

C384(128)

43.4±14.6

Brainglioma

Coffee:T3

vs.T1

Tea:T3

vs.T1

Teaplus

coffee:T3

vs.T1

Coffee:0.09

(0.03–0.24)

Tea:0.33

(0.13–0.86)

Teaplus

coffee:

0.35

(0.15–0.83)

Adjustedfore

nergyintake,physicalactivity,

family

historyof

cancers,family

historyof

glioma,maritalstatus,education,high

risk

occupatio

n,high

-risk

resid

entialarea,

duratio

nof

cellph

oneuse,supp

lement

use,historyof

expo

sure

totheradiog

raph

icX-ray,historyof

head

trauma,historyof

allergy,historyof

hypertensio

n,sm

oking

status,exposureto

chem

icals,drug

use,

person

alhairdye,frequ

entfried

food

intake,frequ

entuseof

barbecue,canne

dfood

sandmicrowave,meatsand

processedmeats,leg

umes

andnu

ts,fruits,

saltandinteractioneffectsof

teaand

coffeeconsum

ption,andBM

I.

Michaud

etal.2010

France,Italy,Spain,

Den

mark,Germany,

Greece,Nethe

rland

s,Norway,Swed

en,U

K

Coh

ort

521,448(343)

25–70

Brainglioma

Coffee:Q5vs.Q

1Tea:Q4vs.Q

1Teaplus

coffee:Q5vs.Q

1

Coffee:0.98

(0.67–1.41)

Tea:1.05

(0.75–1.48)

Teaplus

coffee:

1.02

(0.72–1.44)

Adjustedforage,sex,coun

try,bo

dymass

inde

x,sm

okingstatus,and

education.

Nelsonet

al.2012

USA

Coh

ort

8006

(9)

45–68

Brainglioma

Coffee:≥4cups/day

vs.non

eCoffee:0.89

(0.08–10.02)

Adjustedforage,ed

ucation,andtricep

s

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 3 of 9

Table

1Characteristicsof

theinclud

edstud

ieson

teaandcoffeeconsum

ptionandbraincancer

risk(Con

tinued)

Stud

y,year

Cou

ntry

Stud

yde

sign

Participants

(cases)

Age

(years)

Outcome

Expo

sure

andcatego

ryRR

(95%

CI)

Adjustm

entforcovariates

Tea:≥4cups/day

vs.non

eTea:1.21

(0.22–6.76)

skinfold

thickness

Ogawaet

al.2016

Japan

Coh

ort

106,324(157)

40–69

Braincancer

Brainglioma

Coffee:≥3cups/day

vs.≤

4days/w

eek

Tea:≥3cups/day

vs.≤

4days/w

eek

Braincancer:

Coffee:0.48

(0.23–1.00)

Tea:0.55

(0.17–1.84)

Brainglioma:

Coffee:1.07

(0.70–1.62)

Tea:1.05

(0.54–2.05)

Adjustedforage,sex,BM

I,pack-yearsof

cigarettes

(never

andpast,0–20,>20),

alcoho

lintake(non

andpastand1–3

times/m

onth,d

rinker≤150,150gof

ethano

lper

week),g

reen

tea(≤

4days/

week,1–2cups/day,≥

3cups/day),and

pasthistoryof

allergy,pasthistoryof

diabetes

mellitus.

Abb

reviations:R

Rrelativ

erisk,CI

confiden

ceinterval,P

BCCpo

pulatio

n-ba

sedcase-con

trol

stud

ies,HBC

Cho

spita

l-based

case-con

trol

stud

ies,T3

tertile

3,T1

tertile

1,Q1qu

artile1,

Q4qu

artile4,

Q5qu

artile5

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 4 of 9

location, we found a statistically significant protective ef-fect of tea consumption and the risk of brain cancer inAmerican populations (RR = 0.798, 95% CI = 0.646–0.986), instead of other populations. Detailed results areshown in Table 2.Based on Egger’s test (P = 0.735) and funnel plot

(Additional file 2: Figure S2), there existed no publi-cation bias between tea consumption and brain can-cer risk. Sensitivity analysis was done and the pooledRR ranged from 0.863 (95% CI = 0.699–1.067) to0.947 (95% CI = 0.763–1.176).

Coffee plus tea consumption and the risk of brain cancerFour studies [25, 29–31] were performed to assess theassociation between coffee plus tea consumption and therisk of brain cancer. Figure 4 showed that the RR was0.684 (95% CI = 0.481–0.975) when compared the high-est versus the lowest category consumption of coffeeplus tea.

DiscussionThis meta-analysis included 11 publications involving2583 cases and 1,684,262 participants to assess the

association of coffee consumption, tea consumption, andcoffee plus tea consumption on the risk of brain cancer.We also performed subgroup analyses by study designand geographic location to further explore the associ-ation between coffee and tea consumption and the riskof brain cancer.There were some differences between our meta-ana-

lysis and the study by Malerba et al. [18]. First, Malerbaet al. included six studies for coffee consumption, fourstudies for tea consumption, three studies for coffee plustea consumption to assess the risk of adult glioma. How-ever, we included 11 studies for coffee consumption, 8studies for tea consumption, and 4 studies for coffeeplus tea consumption, which was more than Malerba etal. Second, the results in our analysis suggested the high-est versus the lowest categories coffee consumption andcoffee plus tea consumption were inversely associatedwith the risk of brain cancer, which was different fromthe results of Malerba et al. However, we got a consist-ent result for tea consumption and the risk of brain can-cer with them.Tea and coffee are rich in polyphenols including phen-

olic acids and flavonoids [34], which are known for their

Fig. 1 Flow chart of meta-analysis for exclusion/inclusion of studies

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 5 of 9

antioxidant activity, regulation of heterogeneous metab-olite enzymes, and inhibition of tumor promotion [35],and have been shown to prevent cancer [36, 37]. Fur-thermore, diterpenes and caffeic acid in coffee couldprotect against cancer [38]. However, different tea con-sumed (such as green tea or black tea) and different cof-fee brewing method may vary on the effect of cancer[39]. We did not analyze the detailed effect about the ac-count type of tea consumed and coffee brewing methodon brain cancer due to the limited information in the in-dividual study. Therefore, further researches are war-ranted to explore the potential association about them.

Sensitivity analysis for coffee and tea consumption andthe risk of brain cancer was performed. For coffee con-sumption and brain cancer risk, the association changedfrom being statistically significant to not statistically sig-nificant when removed the study by Malmir et al. 2017(RR = 0.905, 95% CI = 0.754–1.088). Therefore, the resultfor coffee consumption and the risk of brain cancer wasnot stable. Future studies with large participants are war-ranted to further confirm this association. For tea con-sumption and the risk of brain cancer, the association wasnot significant when removed a single study one by one,with pooled RR ranged from 0.863 (95% CI = 0.699–1.067)

Fig. 2 The forest plot of the relationship between coffee consumption and the risk of brain cancer

Table 2 Summary risk estimates of the association between tea and coffee consumption and brain cancer risk

Subgroups Tea consumption (highest vs. lowest category) Coffee consumption (highest vs. lowest category)

Studies, n RR (95% CI) I2 (%) Pheterogeneity Studies, n RR (95% CI) I2 (%) Pheterogeneity

All studies 8 0.897 (0.739–1.088) 29.9 0.189 11 0.785 (0.580–0.984) 65.6 0.001

Glioma 7 0.846 (0.683–1.047) 24.6 0.241 10 0.760 (0.548–0.972) 63.9 0.003

Cohort 6 0.891 (0.755–1.051) 0.0 0.471 8 0.858 (0.700–0.992) 20.1 0.270

Case-control 2 0.658 (0.199–2.179) 81.5 0.020 3 0.507 (0.142–1.810) 90.1 < 0.001

America 4 0.798 (0.646–0.986) 0.0 0.595 6 0.912 (0.740–1.124) 0.0 0.520

Europe 2 1.080 (0.828–1.410) 0.0 0.792 2 1.070 (0.793–1.444) 0.0 0.433

Asia 2 0.643 (0.205–2.016) 79.9 0.026 2 0.217 (0.042–0.896) 84.9 0.010

Oceania – – – – 1 – – –

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 6 of 9

Fig. 3 The forest plot of the relationship between tea consumption and brain cancer risk

Fig. 4 The forest plot of the relationship between coffee plus tea consumption and brain cancer risk

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 7 of 9

to 0.947 (95% CI = 0.763–1.176). For coffee plus tea con-sumption and the risk of brain cancer, the pooled RRchanged to 0.754 (95% CI = 0.543–1.045) when removedthe study by Malmir et al. 2017. Thus, more relevant stud-ies about coffee plus tea consumption and the risk of braincancer are needed.Some limitations existed in our meta-analysis. First,

high between-study heterogeneity existed in some re-sults. Although meta-regression analysis was performedto explore the sources of this high between-study het-erogeneity, we could not find the detailed sources of het-erogeneity due to stratified analysis and meta-regressionanalysis concerning the relationship between coffee con-sumption and the risk of brain cancer. On the otherhand, we used a random-effects model in all the results,which had a wider CI than the fixed-effects model. Thus,random-effects could obtain a more conservative result.Second, 3 of the 11 studies were case-control design,which could lead to some recalling bias, selective bias,and so on. However, case-control studies could also ex-plain the relationship between coffee and tea consump-tion and the risk of brain cancer. Third, we could notperform the subgroup analysis by sex or lifestyle due tothe limited information in each individual study. Consid-ering sex and lifestyle may be a risk factor for coffeeand/or tea consumption, future studies with detailedinformation about sex and lifestyle are warranted tofurther explore the association between coffee or teaconsumption and risk of brain cancer. Fourth, therange of years in coffee or tea consumption may playa different role in the risk of brain cancer due to can-cer maybe consists of various stages during aging[40]. However, the information about the range ofyears about coffee and tea consumption in each indi-vidual study was limited.

ConclusionsFindings from this study suggested that higher consump-tion of coffee may contribute to the lower developmentof brain cancer in Asian populations. Tea consumptionhad an inverse association for the risk of brain cancer inAmerican populations, instead of other populations. Assome limitations existed in our study, future studies withdetailed information about sex, lifestyle, and some otherrelated factors are warranted to further explore the asso-ciation between coffee or tea consumption and risk ofbrain cancer.

Additional files

Additional file 1: Figure S1. Funnel plot for the analysis of publicationbias between coffee consumption and brain cancer risk. (TIF 40 kb)

Additional file 2: Figure S2. Funnel plot for the analysis of publicationbias between tea consumption and brain cancer risk. (TIF 39 kb)

AbbreviationsCI: Confidence interval; OR: Odds ratio; RR: Relative risk

AcknowledgementsNone

FundingNone.

Availability of data and materialsThe tables and figures supporting the conclusions of this article are includedwithin the article.

Author’s contributionsYS conceived and designed the study. YS, ZYW, and YYJ participated in datacollecting. JG analyzed data. YS and ZYW commented on drafts of the paper.All authors read and approved the final manuscript.

Ethics approval and consent to participateNot applicable.

Consent for publicationAll authors consented for publication.

Competing interestsThe authors declare that they have no competing interests.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims in publishedmaps and institutional affiliations.

Author details1Department of Neurology, Tianjin First Center Hospital, No. 24, Fukang Road,Nankai District, Tianjin 300192, China. 2Department of General Surgery,Tianjin First Center Hospital, Tianjin 300192, China.

Received: 17 January 2019 Accepted: 7 March 2019

References1. Li WF, Niu C, Shakir TM, et al. An evidence-based approach to assess the

accuracy of intravoxel incoherent motion imaging for the grading of braintumors. Medicine (Baltimore). 2018;97:e13217.

2. Daumas-Duport C, Scheithauer B, O'Fallon J, et al. Grading of astrocytomas.A simple and reproducible method. Cancer. 1988;62:2152–65.

3. Li Y. Association between fruit and vegetable intake and risk for glioma: ameta-analysis. Nutrition. 2014;30:1272–8.

4. Lian W, Wang R, Xing B, et al. Fish intake and the risk of brain tumor: ameta-analysis with systematic review. Nutr J. 2017;16:1.

5. Lv W, Zhong X, Xu L, et al. Association between dietary vitamin a intakeand the risk of glioma: evidence from a meta-analysis. Nutrients. 2015;7:8897–904.

6. Zhou S, Wang X, Tan Y, et al. Association between vitamin C intakeand glioma risk: evidence from a meta-analysis. Neuroepidemiology.2015;44:39–44.

7. Kyritsis AP, Bondy MLand Levin VA. Modulation of glioma risk andprogression by dietary nutrients and antiinflammatory agents. Nutr Cancer.2011;63:174–84.

8. Horisaki K, Takahashi K, Ito H, et al. A dose-response meta-analysis of coffeeconsumption and colorectal Cancer risk in the Japanese population:application of a cubic-spline model. J Epidemiol. 2018;28:503–9.

9. Lukic M, Guha N, Licaj I, et al. Coffee drinking and the risk of endometrialCancer: an updated meta-analysis of observational studies. Nutr Cancer.2018;70:513–28.

10. Lafranconi A, Micek A, Galvano F, et al. Coffee decreases the risk ofendometrial Cancer: a dose-response meta-analysis of prospective cohortstudies. Nutrients. 2017;9. https://doi.org/10.3390/nu9111223.

11. Ni CX, Gong H, Liu Y, et al. Green tea consumption and the risk of liverCancer: a meta-analysis. Nutr Cancer. 2017;69:211–20.

12. Huang YQ, Lu X, Min H, et al. Green tea and liver cancer risk: a meta-analysis ofprospective cohort studies in Asian populations. Nutrition. 2016;32:3–8.

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 8 of 9

13. Zhan X, Wang J, Pan S, et al. Tea consumption and the risk of ovariancancer: a meta-analysis of epidemiological studies. Oncotarget. 2017;8:37796–806.

14. Huang Y, Chen H, Zhou L, et al. Association between green tea intake andrisk of gastric cancer: a systematic review and dose-response meta-analysisof observational studies. Public Health Nutr. 2017;20:3183–92.

15. Zhou H, Wu W, Wang F, et al. Tea consumption is associated withdecreased risk of oral cancer: a comprehensive and dose-response meta-analysis based on 14 case-control studies (MOOSE compliant). Medicine(Baltimore). 2018;97:e13611.

16. Wang W, Yang Y, Zhang W, et al. Association of tea consumption and therisk of oral cancer: a meta-analysis. Oral Oncol. 2014;50:276–81.

17. Zhou Q, Li H, Zhou JG, et al. Green tea, black tea consumption and risk ofendometrial cancer: a systematic review and meta-analysis. Arch GynecolObstet. 2016;293:143–55.

18. Malerba S, Galeone C, Pelucchi C, et al. A meta-analysis of coffee and teaconsumption and the risk of glioma in adults. Cancer Causes Control. 2013;24:267–76.

19. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ. 2003;327:557–60.

20. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis.Stat Med. 2002;21:1539–58.

21. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detectedby a simple, graphical test. BMJ. 1997;315:629–34.

22. Begg CBand Mazumdar M. Operating characteristics of a rank correlationtest for publication bias. Biometrics. 1994;50:1088–101.

23. Baglietto L, Giles GG, English DR, et al. Alcohol consumption and risk ofglioblastoma; evidence from the Melbourne collaborative cohort study. Int JCancer. 2011;128:1929–34.

24. Burch JD, Craib KJ, Choi BC, et al. An exploratory case-control study of braintumors in adults. J Natl Cancer Inst. 1987;78:601–9.

25. Dubrow R, Darefsky AS, Freedman ND, et al. Coffee, tea, soda, and caffeineintake in relation to risk of adult glioma in the NIH-AARP diet and healthstudy. Cancer Causes Control. 2012;23:757–68.

26. Efird JT, Friedman GD, Sidney S, et al. The risk for malignant primary adult-onset glioma in a large, multiethnic, managed-care cohort: cigarettesmoking and other lifestyle behaviors. J Neuro-Oncol. 2004;68:57–69.

27. Hashibe M, Galeone C, Buys SS, et al. Coffee, tea, caffeine intake, and therisk of cancer in the PLCO cohort. Br J Cancer. 2015;113:809–16.

28. Hochberg F, Toniolo P, Cole P, et al. Nonoccupational risk indicators ofglioblastoma in adults. J Neuro-Oncol. 1990;8:55–60.

29. Holick CN, Smith SG, Giovannucci E, et al. Coffee, tea, caffeine intake, andrisk of adult glioma in three prospective cohort studies. Cancer EpidemiolBiomark Prev. 2010;19:39–47.

30. Malmir H, Shayanfar M, Mohammad-Shirazi M, et al. Tea and coffeeconsumption in relation to glioma: a case-control study. Eur J Nutr. 2017.[Epub ahead of print].

31. Michaud DS, Gallo V, Schlehofer B, et al. Coffee and tea intake and risk ofbrain tumors in the European prospective investigation into Cancer andnutrition (EPIC) cohort study. Am J Clin Nutr. 2010;92:1145–50.

32. Nelson JS, Burchfiel CM, Fekedulegn D, et al. Potential risk factors forincident glioblastoma multiforme: the Honolulu heart program andHonolulu-Asia aging study. J Neuro-Oncol. 2012;109:315–21.

33. Ogawa T, Sawada N, Iwasaki M, et al. Coffee and green tea consumption in relationto brain tumor risk in a Japanese population. Int J Cancer. 2016;139:2714–21.

34. Scalbert A, Williamson G. Dietary intake and bioavailability of polyphenols.J Nutr. 2000;130:2073S–85S.

35. Frei B, Higdon JV. Antioxidant activity of tea polyphenols in vivo: evidencefrom animal studies. J Nutr. 2003;133:3275S–84S.

36. Yang CS, Maliakal P, Meng X. Inhibition of carcinogenesis by tea. Annu RevPharmacol Toxicol. 2002;42:25–54.

37. Ju J, Lu G, Lambert JD, et al. Inhibition of carcinogenesis by teaconstituents. Semin Cancer Biol. 2007;17:395–402.

38. Cavin C, Holzhaeuser D, Scharf G, et al. Cafestol and kahweol, two coffee specificditerpenes with anticarcinogenic activity. Food Chem Toxicol. 2002;40:1155–63.

39. Sachinidis A, Seul C, Seewald S, et al. Green tea compounds inhibit tyrosinephosphorylation of PDGF beta-receptor and transformation of A172 humanglioblastoma. FEBS Lett. 2000;471:51–5.

40. Baker DJ, Peleg S. Biphasic modeling of mitochondrial metabolismdysregulation during aging. Trends Biochem Sci. 2017;42:702–11.

Song et al. World Journal of Surgical Oncology (2019) 17:51 Page 9 of 9